Warmer survey starts were linked to fewer recorded bird taxa in North Carolina in 2005-2024

We tested the link between survey-start temperature and recorded bird taxa in California, Washington, New York, and North Carolina. We used 8,177 pairs of accepted Breeding Bird Survey runs from 1985-2024. Each pair used the same route and observer in consecutive years. We fit separate models for two periods and adjusted for year, wind, sky, date, survey time, duration, and assistant status. In North Carolina in 2005-2024, a 5 C warmer start was linked to 0.75 fewer recorded taxa. The approximate interval, with adjustment for eight tests, was -1.34 to -0.15 taxa. The negative estimate also held in June-only, low-wind, and experienced-observer checks. In New York in 2005-2024, adjustment changed the estimate from -0.44 to +0.04 taxa. The adjusted interval included zero. No common negative link was clear across all four states and both periods. Counts include some non-species taxa. Survey selection, changing bird activity, detection, habitat, and taxonomy limit interpretation. This study does not test air pollution, climate trends, true heat extremes, or population loss. The package contains code, source hashes, input records, pair data, model results, and a figure.

birds · survey weather · observation bias · temperature · recorded taxa · temporal replication · Breeding Bird Survey